Doxis AI.dp, formerly Klippa DocHorizon, is a mature European IDP platform for enterprises that need modular document automation, mobile capture, anonymization, and managed workflows. Suparse is a better fit when you want self-service document extraction, transparent page-based pricing, Python and TypeScript SDKs, custom schemas, and export-ready data without a sales-led rollout.
| Criteria | Suparse | Klippa DocHorizon / Doxis AI.dp |
|---|---|---|
| Best for | Finance, operations, logistics, and product teams that want fast self-service extraction for invoices, receipts, bank statements, tax forms, shipping documents, and custom documents. | Enterprises that need a broader modular IDP platform with mobile scanning SDKs, anonymization, validation, classification, splitting, and managed workflow implementation. |
| Pricing model | Published monthly page-based plans start at $11/month for 100 pages, with no required setup fee. | The current AI.dp page points buyers toward tailored quotes, with pricing depending on volume, workflow complexity, and subscription plan. |
| Free trial | 50 free page credits, no credit card required. | The main product page is quote-led; an official webinar page references €25 of free pay-as-you-go credit, so current trial access should be confirmed. |
| Time to first extraction | Under 60 seconds for supported document types, with AI-assisted schema creation for custom documents. | Can be fast for API-led enterprise integrations, but setup is typically more consultative and module-dependent. |
| SDK | REST API plus Python and TypeScript SDKs. | REST API, a public Node/TypeScript SDK, and native mobile SDKs are documented. Public Python SDK coverage was not found. |
| Schema flexibility | Create and edit custom extraction schemas in the UI without labeled training sets. | Supports ready-made models, presets, field selection, and custom model training, but custom use cases may require configuration or training. |
| Template requirements | Template-free zero-shot extraction with pre-trained schemas and custom schema generation. | Offers ready-made models and configurable workflows; custom document types may require setup or training rather than pure zero-shot extraction. |
| Deployment model | Cloud SaaS by default, with VPC or on-premise option for controlled deployments. | Can be used standalone or as part of a broader Doxis deployment; exact hosting and enterprise packaging should be confirmed during procurement. |
| Setup complexity | Self-service signup, online purchase, and no required implementation services. | More enterprise-oriented, especially when using Doxis deployment options, anonymization, mobile capture, custom training, or broader workflow modules. |
| Review workflow | Built-in side-by-side review, update-in-place editing, validation rules, and audit trail. | The public tutorial flow shows extracted data reviewed against the source document in the UI. |
| Export options | Excel, CSV, JSON, QBO, IIF, Google Sheets, and Xero-compatible CSV. | Public listings describe JSON, XML, CSV, XLSX, UBL, and PDF outputs. |
| Team access | Unlimited user seats with every subscription. | Enterprise access, roles, and SLA terms should be confirmed in the quoted plan. |
Use this table to compare operating model, implementation effort, and workflow depth. Verify quoted-plan details directly before procurement.
| Capability | Suparse | Klippa DocHorizon / Doxis AI.dp |
|---|---|---|
| Structured JSON output | Suparse returns structured JSON through the API and supports spreadsheet exports for operations teams. | Public listings describe JSON output alongside XML, CSV, XLSX, UBL, and PDF exports. |
| Invoice and receipt presets | Pre-trained models extract vendor details, dates, line items, tax, subtotals, and totals without setup. | Official/public sources describe support for 50+ document types, including invoices and receipts. |
| Bank statement support | Suparse extracts transactions and can validate opening balance plus transactions against closing balance. | Reviewed sources describe configurable financial document extraction, but a built-in bank statement reconciliation check is not clearly documented. |
| Tax form and financial document support | Suparse supports invoices, receipts, bank statements, checks, tax forms, energy bills, and other financial documents. | Official/public sources describe financial documents, identity documents, and business forms among 50+ supported document types. |
| Custom document schemas | AI Schema Generator creates extraction schemas from a sample document, and teams can edit fields in the UI. | Custom fields are available, but review excerpts describe limited self-serve customization and model improvement control. |
| Template-free zero-shot extraction | Suparse is designed for template-free extraction across changing layouts without labeled model training. | AI.dp offers ready-made models and configurable workflows, while custom document types may require setup or model training. |
| No model training required | Use pre-trained schemas or AI-assisted schema creation without preparing labeled sample sets. | Standard models are available, while custom document types may require presets, configuration, or model training. |
| Automatic adaptation to layout changes | Template-free extraction is designed to adapt when vendors move fields or change document layouts. | Reviewer excerpts describe limited visible improvement from feedback; test layout-change behavior on your own documents. |
| Handwriting recognition in 100+ languages | Suparse supports handwriting recognition across 100+ languages, including non-Latin scripts. | Public sources describe broad language support, but the exact current language count is not clearly documented. |
| Scanned and degraded document support | Suparse handles scans, mobile photos, skew, blur, poor lighting, faxes, and low-quality images. | Klippa DocHorizon is positioned as an OCR and IDP platform for scanned and digital document processing. |
| Complex table extraction | Suparse handles borderless, nested, variable-row, and multi-page tables. | Table and line-item extraction are part of IDP workflows, but complex multi-page table behavior should be tested on buyer documents. |
| Automatic document classification | Suparse can classify uploaded documents and route them to the correct extraction schema. | Official/public sources identify classification and sorting as core AI.dp capabilities. |
| Automatic multi-document PDF splitting | Suparse can split mixed PDFs into individual records and classify each document. | Official/public sources describe document splitting, and the public JavaScript SDK includes split utilities. |
| Long document auto-chunking | Auto-chunking supports long, multi-page documents without manual splitting. | Multi-page processing should be checked against quoted plan limits, file-size limits, and document-type configuration. |
| Python SDK | Suparse provides a Python SDK for API-driven extraction workflows. | REST API access is documented, but public Python SDK coverage was not found in the current public docs reviewed. |
| TypeScript SDK | Suparse provides a TypeScript SDK for JavaScript and TypeScript applications. | A public Node/TypeScript SDK is available through GitHub and npm as @klippa/dochorizon-sdk. |
| Mobile SDK | Suparse does not currently position around embedded mobile scanning SDKs. | Official mobile SDK material advertises native iOS/Swift and Android/Kotlin SDKs, plus wrappers for Flutter, Xamarin, React Native, and NativeScript. |
| REST API | REST API supports upload, status polling, result retrieval, and bulk operations. | Public materials reference API access and a Swagger endpoint for DocHorizon. |
| Self-service signup | Users can sign up, use free credits, and buy online. | The current product page is quote-led, though public material also references signup/demo access and €25 of pay-as-you-go credit. |
| Under 60 seconds to first extraction | Upload a supported document or create a schema and start extracting immediately. | API setup may be fast for technical teams, but full platform rollout depends on modules, workflow design, and commercial setup. |
| 50 free page credits with no credit card | Suparse includes 50 free page credits with no credit card required. | An official webinar page references €25 of pay-as-you-go credit, but the main product page is still quote-led. |
| Month-to-month subscription | Monthly billing is available, with lower pricing for yearly commitment. | Current public pricing is custom quoted and depends on volume, workflow complexity, and subscription plan. |
| No setup or implementation fee | No required setup fee. Optional consulting is available if a customer wants help. | Implementation, SLA, support, and module costs should be clarified in the quote. |
| Unlimited user seats | Unlimited seats are included with every Suparse subscription. | User and role limits should be confirmed in the selected enterprise package. |
| Full audit trail for edits and extractions | Suparse tracks extractions and user edits, including who changed what and when. | Human review workflows are described, but exact audit trail scope should be confirmed in the contract. |
| Human review workflow | Review extracted data beside the source document, edit values, and apply validation before export. | Human-in-the-loop review is one of Klippa DocHorizon's stated strengths. |
| Invoice self-consistency checks | Suparse can validate subtotal, tax, total, required fields, dates, currencies, and number formats. | Validation workflows are likely available, but invoice math-check scope should be tested on actual documents. |
| Export to Excel, CSV, JSON, QBO, and IIF | Suparse supports Excel, CSV, JSON, QBO, IIF, Google Sheets, and Xero-compatible CSV. | Reviewed sources list JSON, XML, CSV, XLSX, UBL, and PDF, but not QBO or IIF. |
| Direct accounting integrations | Suparse exports accounting-ready files, but does not currently provide native one-click accounting pushes. | Reviewed sources suggest custom integrations rather than native QuickBooks or Xero pushes. |
| Fraud detection and document tampering detection | Suparse does not currently provide built-in fraud or tamper detection. | Older and competitor sources discuss fraud detection, but current official AI.dp pages reviewed did not clearly document the exact fraud-detection scope. |
| Data anonymization / PII masking | Suparse does not currently include built-in PII redaction. | Reviewed sources describe GDPR-oriented anonymization and PII masking as Klippa capabilities. |
| Cloud SaaS with VPC/on-premise option | Suparse is cloud SaaS by default and offers VPC or on-premise deployment where required. | Doxis says AI.dp can be used standalone or as part of a broader Doxis deployment; exact deployment options should be confirmed directly. |
| EU hosting and GDPR posture | Suparse is GDPR-focused with EU hosting, encryption, user-managed retention, DPA availability, and optional zero retention. | Doxis compliance material states GDPR alignment, ISO 27001 certification, EU hosting, and servers certified for ISO 27001, SOC1, and SOC2. |
| No customer data used for model training | Suparse states customer documents are not used to train public AI models. | Reviewed sources did not show an equivalent current model-training commitment. |
Upload real invoices, receipts, bank statements, logistics documents, or custom forms and inspect the output before committing.
Try Suparse FreeSuparse is designed for teams that want extraction running quickly without buying into a larger enterprise IDP rollout.
from suparse import Suparse
client = Suparse(api_key="SUPARSE_API_KEY")
result = client.documents.extract(
file="invoice.pdf",
schema="invoice"
)
print(result.data)It depends on the workflow. Suparse is better for teams that want transparent pricing, self-service setup, Python and TypeScript SDKs, custom schemas, and export-ready finance data. Klippa DocHorizon, now Doxis AI.dp, is better when mobile scanning, anonymization, Doxis platform alignment, or a managed enterprise IDP rollout are core requirements.
Yes. Official Klippa material now uses Doxis AI.dp (ex-Klippa DocHorizon), and Klippa's about page says Klippa DocHorizon is now Doxis AI.dp after Klippa joined Doxis in March 2025.
Suparse publishes entry pricing starting at $11/month for 100 pages and includes 50 free page credits. The current AI.dp page points buyers toward tailored quotes based on volume, workflow complexity, and subscription plan, so a like-for-like cost comparison requires a Klippa/Doxis quote.
Yes, but it is not a one-click migration. You should map each Klippa output field to a Suparse schema, test representative documents, compare JSON and spreadsheet outputs, and rebuild any downstream integration logic that depends on Klippa-specific response shapes.
There is overlap on invoices, receipts, bank statements, and custom business documents. Public Klippa/Doxis material describes 50+ supported document types, including identity documents and business forms, while Suparse focuses on finance, logistics, HR, healthcare, and custom document extraction with schema-driven outputs.
No. Suparse does not currently provide built-in fraud detection, tamper detection, or PII masking. Klippa/Doxis public material supports anonymization; fraud-detection scope should be verified against current AI.dp documentation and the quoted modules.
Suparse provides Python and TypeScript SDKs plus a REST API. Klippa/Doxis public material supports REST API access, a public Node/TypeScript SDK, and mobile SDKs for iOS, Android, Flutter, Xamarin, React Native, and NativeScript; public Python SDK coverage was not found.
Use real documents to compare extraction quality, review effort, schema fit, and export readiness.
Start Free